This invention relates to statistical methods of eliminating the effects of doublets and carryover from sample to sample in a flow analysis system.
Fluorescently labeled particles are being used in a variety of applications. By varying fluorescent dye concentration and/or the emission wavelengths of the dyes, it is possible to create an almost limitless number of fluorescently distinguishable particles. One such technology currently available uses polystyrene microspheres into which are incorporated precisely controlled quantities of two fluorescent dyes. See, for example, U.S. Pat. No. 5,981,180, granted to Chandler et al.
Once separately addressable microsphere sets have been constructed, different specific biomolecular reactions, including but not limited to, DNA, immunoassay, receptor-ligand, or enzyme-based assays, are performed on the discrete microsphere sets. When passed through a flow analyzer, the reaction can be classified by the internal microsphere fluorescent signature (i.e., by the separate spectral addresses), while the biomolecular reaction is measured using fluorescence of a different color (wavelength), i.e., the reporter signal.
However, any time laboratory samples are performed sequentially, regardless of instrument type or methodology, carryover from the previous sample(s) is always a concern. In the past, elaborate wash cycles have been performed in an attempt to solve this problem. The drawback to this xe2x80x9csolution,xe2x80x9d however, is increased cycle time, and there is no guarantee of 100% elimination of carryover.
The present invention provides for substantially 100% elimination of carryover effects and doublets by taking advantage of the flow analyzer""s capability to accurately distinguish the separately addressable microsphere sets and to precisely measure the biomolecular reaction that has occurred on the surface of the microspheres. Also, the number of microspheres measured per analyte ranges over a wide number, preferably between 1 and about 10,000. Typically, a minimum of about 100 microspheres per analyte can be analyzed to provide a reliable result. The amount or intensity of surface (or reporter) fluorescence is averaged among the about 100 pieces of data to measure the extent of the biomolecular reaction. Coefficients of variation (CV""s), mean, and standard deviation can be calculated for the measured value.
The maximum number, say N, of microspheres that could ever carryover from sample to sample is physically measured and/or statistically calculated. According to a particular embodiment of the present invention, 2N+100 microspheres could be analyzed for each biomolecular reaction, yielding 2N+100 pieces of measurement data. The analysis package can then be tailored to discard N highest measured values and N lowest measured values, and average the extent of the biomolecular reaction using only the remaining 100 pieces of measurement data.
According to a preferred embodiment of the invention, only microspheres carrying surface fluorescence values plus or minus two standard deviations from the mean value would be taken into account towards the ultimate assay measurement. Microspheres from a prior sample can be expected to carry surface fluorescence values falling either inside or outside this window.
If the fluorescence values of microspheres from a previous sample fall outside this window, their contribution to the final measurement value would be substantially eliminated or minimized because enough microspheres would have been measured to account for at least about twice the maximum possible number of carryover microspheres. It should now be apparent that if only, say, a total of N microspheres would have been measured initially, a large percentage of the total could be have due to carryover microspheres. The resulting CV""s, mean, and standard deviation values would then have been greatly influenced by the surface fluorescence values of such carryover microspheres. The ultimate reading, then, would have been erroneous.
Of course, if the carryover values fall inside the window, these carryover values would be taken into account; however, these carryover values would not change the result for the particular sample of interest.
In other words, carryover effects that could have altered the biomolecular result would have been eliminated. In addition, any doublets, which would have displayed approximately twice the surface fluorescence values of a single microsphere and, hence, could have erroneously influenced the biomolecular measurement, would have been eliminated.
In much the same manner, the maximum number of particles that could potentially be misclassified (despite the unique spectral address of microspheres in a subset) could be determined. As many xe2x80x9cextraxe2x80x9d microspheres could then be analyzed to negate any possible effect the misclassified microspheres might have on the assay results.
Accordingly, the present invention provides a method of reducing flow analysis error in at least one assay comprising: (a) passing a population of one or more subsets of spectrally addressable microspheres, which had been exposed to a sample suspected of harboring one or more analytes of interest, through a flow analyzer, each subset of spectrally addressable microspheres being distinguishable by at least one characteristic from at least one other subset of spectrally addressable microspheres, said microspheres exhibiting a surface fluorescence value indicative of a biomolecular reaction having taken place on the surface of said microspheres in the presence of said one or more analytes of interest; (b) discarding a number of highest surface fluorescence values and a number of lowest surface fluorescence values, which number equals a maximum number of microspheres, N, carried over from a prior assay; and (c) calculating the extent of the biomolecular reaction, if any, based on the remaining, undiscarded surface fluorescence values. In preferred embodiments of the invention, the method further comprises determining at least the coefficients of variation, mean, and standard deviation of the surface fluorescence values. What is more the calculation of the extent of the biomolecular reaction is carried out only one those remaining, undiscarded surface fluorescence values that are within two standard deviations of the mean.
Consistent with the objectives of the present invention, the number N is predetermined or estimated. An approximate minimum number, M, of microspheres required to afford a reliable assay can also be determined or estimated. Preferably, in the assay of the present invention at least about 2N+M microspheres are passed through the flow analyzer, more preferably at least about 2(N+M). Most preferred assays are those which comprise a multiplexed assay, in which multiple analytes are determined substantially simultaneously.
In another object of the invention a method is provided of reducing microsphere misclassification error in at least one assay comprising the steps of: (a) passing a population of one or more subsets of spectrally addressable microspheres through a flow analyzer, each subset of spectrally addressable microspheres being distinguishable by at least one characteristic from at least one other subset of spectrally addressable microspheres, said at least one characteristic expressed as a certain value measurable by the flow analyzer; (b) discarding a number of highest values and a number of lowest values, which number equals a maximum number of microspheres, Nxe2x80x2, which can be misclassified; and (c) determining the classification of the one or more subsets of microspheres based on the remaining, undiscarded values. The source of microspheres, which can be misclassified, comprises microspheres carried over from one or more prior assays or comprises microspheres present in the at least one assay.